Supermarket Markov Chain Simulation

by Statistically Significant Four

A Monte Carlo Markov Chain simulation to predict the customers' behavior in a hypothetical supermarket with below layout:

Usage

$ python supermarket.py

Repo Structure

1. MCMC Simulation

• supermarket.py -> the executable script that runs the simulation
• proba.py -> the calculation of the transition matrix
• visualization.py -> for an old-fashioned videogame-like visualization of the Monte Carlo Simulation
• output -> folder with the simulation output data

2. Notebooks

• EDA.ipynb -> exploratory data analysis on the raw data
• theory.ipynb -> summary on Markov-Chains theory

3. Data

• data -> The daily supermarket raw data that form the dataset of this model (in csv format)
• data/cleaned-up -> crean data created in EDA.ipynb, and ready for calucating the transition matrix

4. Output

• MCMC_Simulation -> The results of the simulation in .csv file after running supermarket.py
• dungeon_supermarket map -> buited in visualization.py

Credits

This project was done in collaboration with Francesco Mari, Behzad Azarhoushang & Vlasis Tritakis.